The energy consumption by non-residential consumers in China accounts for a significant proportion of the total energy consumption in the society. Thus, accurate non-residential load forecasting plays an increasingly essential role in the future grid planning and operation. In this paper, a method based on LSTM recurrent neural network is proposed to predict the load of non-residential consumers using multiple correlated sequence information. First, the k-means is employed to analyze the daily load curves of non-residential consumers, classify and mine the consumer’s energy consumption behavior patterns. Then, the Spearman correlation coefficient is applied to investigate the time correlation under multiple time series for non-residential consumers. It is found that there exist multiple related time sequences such as adjacent time points, the same time points in adjacent days, and the same day in adjacent weeks among data samples for a specific consumer. Therefore, a non-residential load forecasting framework based on multiple sequence LSTM recurrent neural networks is presented. The proposed framework is tested on a real data set, which contains 48 non-residential consumers’ energy consumption data in China, and outperforms the other load forecasting approaches. Experiment results show that this method can effectively utilize the multiple sequence information, and successfully capture the dependencies among these sequences.

As a networked cyber physical system, smart grid plays an important role in improving energy efficiency, which has higher prediction accuracy requirement for load forecasting to realize intelligent dispatching. The existing works of load forecasting have considered the external factors to improve prediction accuracy, but ignored their varying and different impacts in different time. In order to solve this problem, firstly, a new method is designed to select external meteorological factors and construct new features to deal with the different impacts of external factors in different seasons. Secondly, a fuzzy processing method of calendar factor is proposed to explore the varying impact of the calendar factor in the holiday period. Finally, our proposed method is combined with LSTM and other three typical machine learning models, and experiments are performed on a load data set from the real power system. The results show that our proposed method can improve the prediction accuracy of the typical machine learning based models. Especially for LSTM, the MAPEs reach to 3.8% and 3.1% in winter and summer, with the improvements of 0.6% and 0.3% respectively.

With the increase of grid-connected photovoltaic (PV) power plants, measuring the uncertainty of PV output forecast due to multiple meteorological factors has become an important method to ensure the stable operation of the power system. However, the manual similar days' extraction methods of historical power features and cloud motion features, which are commonly used for PV spatial-temporal prediction, are inefficient and coarse-grained. In this paper, focusing on automatic attention of the local similar sequences in daily features, we improve the attention scoring method of the original transformer to increase the calculation accuracy of the similarity of PV power sequences. We construct an end-to-end PV power prediction framework using an encoder-decoder architecture. Based on the spatial-temporal similarity characteristics of PV output, the fine-grained temporal and cloud-covered spatial attentions embedded in the prediction framework are proposed using the improved similar sequence attention module. Among them, fine-grained temporal attention can dynamically position and utilize similar historical PV power sequences for the predicted period; cloud-coverage spatial attention automatically captures the neighbor site's matching PV power sequences. Experiments were conducted on a real PV dataset, and the root mean square error was reduced by 21.34% on average compared to the baseline model. The results proved the effectiveness of the proposed model.

Unmanned Aerial Vehicle (UAV) inspection has gradually replaced manual inspection of transmission tower, which produces many images. While it is laborious and time-consuming to manually analyze these images, there are also challenges in automatically detecting small-size defective components such as bolts in transmission tower images, due to problems including complex background, small size, and many similar objects of bolts. In this paper, by virtue of multi-scale features and context information, we propose a deep neural network named Camp-Net (Context Information and Multi-Scale Pyramid Network) to identify bolts defect in transmission tower images. First, multi-scale feature fusion combines deep features and shallow features in convolutional networks to detect small-size bolts. Second, context information fusion puts the information around bolts into the detection network to remove the disturbance of complex background and similar objects. An image dataset containing defective bolts and normal bolts is constructed for model training and testing. Experimental results show that bolts with loose pins and bolts without pins among fittings in transmission tower can be accurately identified with the proposed model. The Average Precision (AP) of defective bolts detection of this model can be 11.4% higher than that of the commonly used high performance model, Faster R-CNN.

Long-term time series forecasting (LTSF) provides substantial benefits for numerous real-world applications, whereas places essential demands on the model capacity to capture long-range dependencies. Recent Transformer-based models have significantly improved LTSF performance. It is worth noting that Transformer with the self-attention mechanism was originally proposed to model language sequences whose tokens (i.e., words) are discrete and highly semantic. However, unlike language sequences, most time series are sequential and continuous numeric points. Time steps with temporal redundancy are weakly semantic, and only leveraging time-domain tokens is hard to depict the overall properties of time series (e.g., the overall trend and periodic variations). To address these problems, we propose a novel Transformer-based forecasting model named InParformer with an Interactive Parallel Attention (InPar Attention) mechanism. The InPar Attention is proposed to learn long-range dependencies comprehensively in both frequency and time domains. To improve its learning capacity and efficiency, we further design several mechanisms, including query selection, key-value pair compression, and recombination. Moreover, InParformer is constructed with evolutionary seasonal-trend decomposition modules to enhance intricate temporal pattern extraction. Extensive experiments on six real-world benchmarks show that InParformer outperforms the state-of-the-art forecasting Transformers.

Probabilistic solar irradiance forecasting has received widespread attention in recent years, as it provides more uncertainty information for the future photovoltaic generation. In this study, a hybrid probabilistic solar irradiance prediction method is proposed, which combines a deep recurrent neural network and residual modeling. Specifically, the long short‐term memory‐based point prediction using historical records and related features is applied to obtain deterministic forecasts. Next, these deterministic forecasts are employed as inputs to estimate the residual distributions. Furthermore, maximum likelihood estimation is utilized to compute the parameters of the residual distribution. Finally, the point prediction and residual distribution jointly generate the final probabilistic forecasting results. Compared with other deterministic and probabilistic forecasting models, the proposed method yields promising results on a publicly available dataset. © 2020 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.